IJCST Vol.
1, Issue 2, December 2010
Real Time Expert System - Its Applications
1
Amardeep Singh, 2Monika Verma
Swami Vivekanand Institute of Engineering & Technology, Banur, Punjab, India.
E-mail : 1hodcse@sviet.ac.in, 2monikaverma007@gmail.com
Abstract
This paper is about Real time expert system. Main reason for
using a real-time expert system is to reduce the cognitive load
on users or to enable them to increase their productivity without
the cognitive load on them increasing. Then we discussed the
architecture of real time expert system, life cycle of real time
expert system and main stress is on the application of Real-
Time Expert System.
Keywords
Real time expert system, life cycle, applications.
I. Introduction
Real-time expert systems [1] are on-line knowledge-based
systems that combine analytical process models with
conventional process control while reasoning about the past,
present, and future to assess on-going developments and plan
appropriate actions.
Indicators that a real-time expert system might be appropriate,
especially when conventional techniques have failed or are
impractical, include problem-solving situations where humans
suffer from cognitive overload, fail to effectively monitor
all available information, are unable to resolve conflicting
constraints, are expensive.
Many "real-time" expert systems are 'soft' real-time systems, in
that they claim to be fast. A 'hard' real-time system would have
features that guarantee a response within a fixed amount of
real-time (e.g., bounded computation, not just a fast match-
recognize-act cycle).
Systems like G2 use event-driven processing as a method of
limiting forward chaining.
II. Real time expert system architecture [2]
Here we have passive elements like data files and knowledge
base which are denoted by square boxes and active elements
are data, knowledge manager, processes which are denoted
by dashed ovals. Data exchange between active and passive
elements is shown by arrows and communication between
active elements is shown by dashed arrows. We have divided
this architecture into two subsystems: intelligent subsystem
and real time subsystem. The interface divides and connects
the two subsystems.
Fig.1: Real time expert system architecture
150  International Journal of Computer Science and Technology
ISSN : 2229-4333(Print) | ISSN : 0976-8491(Online)
III. Life cycle of Real time expert system
Lifecycle of real time expert system [3] is similar to that of any
software system. Main stages are shown by rectangular boxes.
Basic sequence of any real time expert system is:
1. Task analysis and specification
2. Software design
3. Coding
4. Testing
5. Documentation and Implementation
Fig. 2: Life cycle of real time expert system
A. Task analysis and specification
As expert system development is very large project and so work
begins by establishing the requirements for all system elements
and then allocating some subset of these requirements to
software. This system view is essential when the system must
interface with other elements such as hardware, people and
other resources. System is the basic and very critical requirement
for the existence of software in any entity. So if the system is
not in place, the system should be engineered and put in place.
Once the ideal system is engineered, the development team
studies the software requirement for the system.
This process is also known as feasibility study. In this phase, the
development team visits the customer and studies their system.
They investigate the need for possible software automation
in the given system. By the end of the feasibility study, the
team furnishes a document that holds the different specific
recommendations for the candidate system. It also includes
the personnel assignments, costs, project schedule, target
dates etc. The requirement gathering process is intensified
and focused specially on system. To understand the nature of
the program to be built, the system engineer or "Analyst" must
understand the information domain for the software, as well
as required function, behavior, performance and interfacing.
The essential purpose of this phase is to find the need and to
define the problem that needs to be solved.
B. Software Design
In this phase, the expert system development process, the
w w w. i j c s t. c o m
ISSN : 2229-4333(Print) | ISSN : 0976-8491(Online)
system's overall structure and its nuances are defined. In terms
of the client/server technology, the number of tiers needed
for the package architecture, the database design, the data
structure design etc. are all defined in this phase. An expert
system development model is thus created. Analysis and
Design are very crucial. Any problem in the design phase could
be very expensive to solve in the later stage of the software
development. Much care is taken during this phase. The logical
system of the product is developed in this phase.
C. Code Generation
The design must be translated into a machine-readable form.
The code generation step performs this task. If the design
is performed in a detailed manner, code generation can be
accomplished without much complication. Programming
tools like compilers, interpreters, debuggers etc. are used to
generate the code. Different high level programming languages
like C, C++, Pascal, and Java are used for coding. With respect
to the type of application, the right programming language is
chosen.
D.Testing
Once the code is generated, the system testing begins. Different
testing methodologies are available to unravel the bugs that
were committed during the previous phases. Different testing
tools and methodologies are already available. Some companies
build their own testing tools that are tailor made for their own
development operations.
E. Documentation and Implementation
The development and maintenance is a staged roll out of the
new application, this involves installation and initial training and
may involve hardware and network upgrades. Expert System will
definitely undergo change once it is delivered to the customer.
Change could be happen because of some unexpected input
values into the system. In addition, the changes in the system
could be directly affecting the software operations. The
software should be developed to accommodate changes that
could happen during the post implementation period.
Some special steps which help in the development of real time
expert system are:
1. Design prototype
2. Special test tools
3. Test plan
1. Design prototype
Its function is to test and demonstrate the operation of a partial
solution, knowledge representation and reasoning technique.
Only a part of the knowledge base and only a few functions
are implemented so this is a downsized partial version of the
final system.
2. Special testing tools
Expert system shells offer special testing tools to test the
stand alone functions of intelligent subsystem. These tools
test the interface and operation of a real time expert system
under different time varying real time conditions. These tools
are needed as it is very difficult to test ant real time system
properly.
3. Testing plan
Since it is very difficult to test the real time system properly so
to prepare a test plan in advance is must. Following situations
w w w. i j c s t. c o m
IJCST Vol. 1, Issue 2, December 2010
should be handled with care:
1. Normal operation with various extreme loads
2. Abnormal and faulty operation modes like missing data,
missing elements, faulty elements, resources not available
etc.
3. Conflict in data and knowledge elements.
IV. Real-Time Expert System- Applications
Real time expert systems [4] are being used in almost all fields
like medical, aerospace, communications, process control,
robotics etc. Following are few expert systems used in medical
field:
1. CAPS: CAPS is a system that combines a pattern-
recognition module with a system module to perform respiratory
and anesthesia monitoring. The pattern-recognition module
characterizes and analyzes each segment of a capnogram.
(A capnogram is a carbon dioxide wave form produced by
monitoring patient respired gases during surgery.) The analysis
of the capnogram is then used by the expert system module
to generate a probable diagnosis and recommend therapy or
equipment adjustments. The system was implemented using
the EXSYS tool on an IBM PC/AT.
2. FORTES: Forth Oriented Real-Time Polysomnographer
Expert System. FORTES was developed at the Stanford University
of Medicine to perform real-time sleep staging. Sleep staging
is a means of classifying a person's state while asleep. It is
usually performed by a trained human scorer who evaluates
a linear time varying graph of the electroencephalograms,
electrocardiogram, and eletromyographic activity while a
person is sleeping. FORTES is implemented on an IBM PC and
is interfaced to a polygraph with a standard digital to analog
converter. The Ventilator Manager VM was one of the first and
most significant applications of knowledge-based systems to a
real-time domain. VM interprets online physiological data in an
intensive care unit. An automatic monitoring system provides
VM with the values of 30 physiological measurements at 2 or
10 minute intervals. The data are used to manage postsurgical
patients receiving mechanical ventilatory assistance. VM
maintains a set of patient-specific expectations and goals
for future evaluation. It is expectation driven and uses the
current and past patient history to establish guidelines for
patient measurements. The guidelines are used to dynamically
establish upper and lower limits for comparison with each new
measurement from the monitoring system. VM responds with
suggestions to clinicians and periodic summaries. VM uses
forward chaining, checks that information previously acquired
is still valid for making conclusions, and cycles through the rule
set each time new information is available.
3. STOCHASM : STOCHASM is a computer program for
performing real time fault detection and diagnosis for the
lubrication oil subsystem in a gas turbine propulsion unit on a
United States Navy surface ship. The system utilizes knowledge
of the order in which alarms are triggered to diagnose the
cause of malfunctions. Malkoff (1986) reports the system
had learning capabilities that allowed it to automatically adapt
to changes in the system and the environment. STOCHASM
was interfaced to a simulator and works in real time. Several
different approaches for performing temporal reasoning.
4. G2 Real Time Expert System is used to monitor complex
industrial processes and to assist in problem identification.
  International Journal of Computer Science and Technology  151
IJCST Vol. 1, Issue 2, December 2010
5. The Emergency Procedures Expert system (EPES) was
designed to be one of the cooperating expert systems that would
make up a Pilot's Associate. EPES works in the domain of in-flight
emergency procedures to detect an emergency, warn the pilot,
and initiate corrective action. The system can be overridden
by the pilot at any time. A prototype of EPES was implemented
in ZetaLisp on a Lisp machine. The knowledge base for EPES
includes parts, goals, and rules and implements emergency
procedures from an F-16 flight manual. EPES interfaces with the
pilot using an expert message display, as well as a simulated
aircraft [5] control system. The Expert Navigator was designed
to monitor, manage, and reconFig. navigation sensors (such
as radio aids, inertial navigation systems, and digital terrain
aids) on board an advanced tactical aircraft.
6. The Heuristic Error Analyzer for Telemetry (HEAT). HEAT
analyzes space shuttle telemetry data that are collected and
down linked to a ground station every second. Once validated,
the telemetry data are then transmitted to NASA mission
controllers. HEAT distinguishes between routine telemetry
errors caused by a loss of signal as the shuttle moves out of
range and a more serious problem requiring human attention.
HEAT was implemented in a Franz Lisp version of OPS5 on a
VAX 11/780 operating under Unix. To accommodate real time
demands of the incoming telemetry status data, a separate C
program was written that translates the telemetry status data
into a Lisp compatible format which is then inserted into a
queue called the AI log file. A separate Lisp program intervenes
to cause the OPS5 code to execute a specified number of rules
and then stop and check the log file for an additional data.
Messages are read either one at a time or in clusters of 20
or more depending on how far HEAT is behind in processing.
Because the shuttle is out of range 12 minutes every hour,
HEAT was also designed to purge unused data when there
is nothing else to do, thus minimizing the impact on memory
usage. Based on trial runs with simulated data, the developers
concluded that OPS5 is not the optimal system architecture for
real-time data processing. One reason given is that the RETE
algorithm used in OPS5 was designed for environments where
the knowledge base changes slowly. In real-time data analysis
problems, the knowledge base changes too rapidly for the RETE
algorithm to be optimal (Skapura and Zoch 1986).
7. The Predictive Monitoring System (PREMON). PREMON
uses the explicit model of a device to perform real-time
monitoring. The system has been tested on the partial model
of the mirror cooling circuit of the Jet Propulsion Laboratory
space simulator. PREMON has three interacting capabilities:
(1) causal simulation to generate predictions about the
behaviour of a physical system, (2) sensor planning to assess
the importance of a device's behavior and allocate sensor
resources appropriately, and (3) sensor interpretation to verify
expected sensor values against actual sensor readings and
raise alarms when necessary (Doyle, Sellers, and Atkinson
1987).
8. The Spacecraft Control Anomaly Resolution Expert
System (SCARES). SCARES attempts to automate the diagnosis
of anomalies in the attitude control system of a spacecraft with
the ultimate goal of achieving spacecraft autonomy. SCARES
has three major components, only one of which performs in real
time. The SCARES monitor performs three real-time checks on
down linked telemetry data: a limit check on individual telemetry
152  International Journal of Computer Science and Technology
ISSN : 2229-4333(Print) | ISSN : 0976-8491(Online)
points, a rate check on two or more telemetry points in the
same channel, and cross-channel checking tests for telemetry
consistency. Hamilton (1986) concludes that SCARES, which
is currently in a prototype stage, shows "how expert system
technology can be applied in large real-time fault diagnostic
systems to detect, diagnose, and recover from complicated
anomalies." SCARES is implemented in Lisp on a Symbolics
3670. SECURE. In the domain of commercial transport aircraft,
SECURE continuously observes the flight environment and
evaluates the situation for possible faults that could threaten
flight safety. Faults requiring further action are detected by
comparing the correct aircraft state with the observed aircraft
state. A recovery procedure is derived and patched into the
current flight, and the crew is informed of the threat and
the recovery procedure if the crew's actions have indicated
that the threat has not been noticed. SECURE's knowledge
base is frame based and is organized according to the flight
contexts of the takeoff phase, flight phase, and landing phase.
Only the knowledge appropriate to the current flight context
is accessed. SECURE was developed using a script-based
approach implemented in MACLISP on a PDP-10 computer
and works well with a "normal" flight containing common
emergencies. The first phase of the planning-based SECURE
has successfully flown a simplified DC-10 aircraft simulation
running on a PDP-10 between any pair of six airports and has
planned recovery actions for several failures. About 230 rules
describe the planning control knowledge (Chen 1985).
9. Expert System for Engineering. This expert system
using fuzzy logic method as an engine to enable this system
operates adaptively. This expert system was developed to
help first year engineering student gain deep understanding
of fundamentals to be able to follow the more advanced topics
in the engineering fields. This ITS will help adaptively adjust the
training for each particular student on the base on his/her pace
of learning. ITS will monitor the student’s progress and have
the ability to make decision about the next step in training. Fig.
below show the architecture of ITS for teaching engineering
student which was embedded expert system inside.   This
expert system using fuzzy rule based decision making system
that would guide the ITS’s  behaviour. For each student, this
expert system will draw the information regarding student’s
performance against the membership function for each topic,
difficulty and importance level.
Fig. 3: Structure of ITS to teach engineering student. Adapted
w w w. i j c s t. c o m
ISSN : 2229-4333(Print) | ISSN : 0976-8491(Online)
from Zorica Nedic, Vladimir Nedic and Jan Machotka (2002)
10. The Environmental Control Expert System in Space
(ECESIS). ECESIS was designed for use on a manned
space station to shift modes for various subsystems of an
environmental control-life support subsystem (EC-LSS) as the
station passes from shadow to sun. ECESIS monitors the EC-
LSS and performs various actions in response to detected
events. The system is implemented in Yet Another Production
System (YAPS) and incorporates both rule based and semantic
net architectures (Dickey and Toussaint 1984).
11. COOKER is a real-time process monitoring and
operator advisory system for batch manufacturing processes.
It runs on a Symbolics 3640 Lisp machine connected by an IBM
AT microcomputer to a Honeywell TDC 2000 process control
system and a programmable logic controller. COOKER has four
main subsystems:
data frames, data gatherer, operator interface, and inference
engine. The latter three subsystems run as concurrent
processes. The data gatherer sends data requests to, and
buffers data from, the IBM AT. The operator interface manages
all windows, displaying advice and questions to the operator,
and receives replies. The inference engine handles un buffering
data from the data gatherer into data frames, receives
replies from the operator, runs the monitoring and problem-
recognition mechanism on the goals, and runs any problem
solving required by the goals. The system uses a knowledge
representation scheme called Goal/Subgoal (GSG). Allard and
Kaemmerer (1987) report that COOKER has been installed in
a manufacturing plant.
V. Conclusion
This paper focuses on architecture of real time expert system.
And its life cycle is almost similar to that of a software
development life cycle (SDLC) which includes various phases
like analysis, designing, coding, testing, implementation and
maintenance. Then various applications in real world have been
discussed and we came to know various fields where real time
expert systems are being used.
References
[1] Ali, M., and Scharnhorst, D. A. 1985. Sensor-Based Fault
Diagnosis in a Flight Expert System.
[2] In Proceedings of the Second Conference on Artificial
Intelligence Applications, 49-52. Washington D.C. IEEE
Computer Society.
[3] Park, J. 1986. Toward the Development of a Real-
Time Expert System. Journal of Forth Applications and
Research.
[4] Schwartz, T. J. 1987. Real Time Processing on the PC.
Spang Robinson Report.
[5] Shaw, R. 1987. RESCU—On-Line Real-Time Artificial
Intelligence. Computer-Aided Engineering Journal
Er. Amardeep Singh is working as AP-CSE.,
S.V.I.E.T., Banur and is also performing the
job of Head of Computer Sc. & Engineering.
and Head of the Department in Swami
Vivekanand Research & Technology Park.
Er.Amardeep Singh has done B.Tech in
Computer Science & Engg. & M.Tech.
w w w. i j c s t. c o m
IJCST Vol. 1, Issue 2, December 2010
in Information Technology from Guru Nanak Dev University
Campus, Amritsar. He has also done Masters in Business
Administration as dual specialization in Marketing & Human
Resource Management. Now he is pursuing Ph.D. in Computer
Sciences. His area of interest is Computer Networks, Parallel
Computation & Distributed Computing. He is the life member
of Indian Society For Technical Education (I.S.T.E.) and has
published & presented papers 05 in International Journals,
07 in International Conferences as well as 18 in National
conferences. He is also serving on the Editorial Board of
International Journal of Computer Science & Technology (IJCST)
and International Journal on Electronics and Communication
Technology (IJECT). He has also written & published 07 highly
acclaimed books. He has attended 10 weeks Short Term/FDP/
Workshop programs at various Institutes. He has Teaching &
Industry Experience of more than 8.
Er. Monika Verma is currently working as an
Assistant Professor in Computer Science
and Engineering Department at Swami
Vivekanand Institute of Engineering &
Technology, Banur. She holds the degree
of M.Tech in Computer Scienece-Hons. and
B.Tech in Computer Science & Engineering.
She was the Topper during M.Tech in her
University. She has around 6 years of
teaching experience. Her research interests include Digital
Image Processing. She has authored and co-authored many
papers in International Journals, International Conferences
and National Conferences and has attended many short term
courses. She is Life Time Member of ISTE. She is also serving
on the Editorial Board of International Journal of Managing
Information Technology (IJMIT) , Signal and Image Processing:
An International Journal (SIPIJ), The International Journal on
Computer Science and Technology (IJCST) and International
Journal on Electronics and Communication Technology
(IJECT).
  International Journal of Computer Science and Technology  153